UC San Diego

Abstract Introduction Obstructive sleep apnea (OSA) is prevalent in the bariatric surgery population and has been associated with increased perioperative risk, especially if OSA is moderate-severe (apnea-hypopnea index ≥ 15/h). Consequently, screening for OSA is recommended as part of the preoperative evaluation. Several screening tools for OSA have been developed; however, some tools lack validation and their relative performance is unclear. The purpose of this study was to compare four existing screening tools (Epworth Sleepiness Scale (ESS), STOP-BANG, NO-OSAS, and No-Apnea) with regards to the ability to identify patients with moderate-severe OSA among bariatric surgery patients. Methods We retrospectively reviewed data from Jan 2015 to Mar 2019 for adult patients presenting consecutively to UC San Diego for first-time bariatric surgery who had undergone a home or in-lab sleep study (within one year of the initial encounter for bariatric surgery), which is our standard of care. We compared the accuracy of the four screening tools for detecting moderate-severe OSA based on the area under the receiver operating characteristic curves (AUC). Subgroup analyses were explored based on sex, BMI, and ethnicity (Hispanic/Latino vs non-Hispanic/Latino). Results Of the 214 patients (83.2% female, median age 39 years) included in the study, 45.4% had moderate-severe OSA. STOP-BANG (AUC 0.75 [95%CI: 0.68 to 0.81]) and NO-OSAS (AUC 0.76 [95%CI: 0.69 to 0.82]) had similar performance (p 0.62); both performed significantly better than the ESS (AUC 0.61 [95%CI: 0.54 to 0.68]; p 0.02 for both). STOP-BANG and NO-OSAS tended to perform better in the female vs male subgroup, but this finding did not reach statistical significance. Conclusion STOP-BANG and NO-OSAS are superior to the ESS when screening bariatric surgery patients for moderate-severe OSA. In future analyses we will further explore if adjustments of standard cut-offs improve test characteristics (i.e. sensitivity/specificity) when screening bariatric surgery patients (analyses ongoing). Support None.